Does the Couch Need to Be Included for Accurate IMRT Planning?
M Gulam*, A Gopal, K Chin, J Kim, A Orfali, J Gordon, I Chetty, T Nurushev, Henry Ford Health System, Detroit, MISU-E-T-386 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Purpose: To evaluate the effect of using couch structures in IMRT planning
Methods: Three TG-119 IMRT plans (Head&Neck, Prostate, C-shape hard) developed for commissioning the Varian Eclipse AAA 10.0.28 algorithm were delivered on a Varian 21EX with Exact Couch. This algorithm allows accounting for the effects of the treatment couch via a separate structure outside the external contour. Each plan was calculated twice (with and without couch), keeping monitor units and leaf motions the same to facilitate comparison. Plans were delivered to an ion chamber and an EBT2 film in a 30x30x20 cm3 phantom at two locations reflecting high and low dose regions per TG-119. Additionally, 15 clinical cases (five each of 6MV H&N, 6MV prostate and 18MV prostate were delivered; four of the ten prostate cases included pelvic lymph node involvement. The film analysis (using in-house software) required >95% points to pass a gamma criterion of 3%,3mm.
Results: For the TG-119 plans 99.1%±2.7% and 94.6%±19.9% of the points met the gamma 3%,3mm with and without the couch model respectively; the major contribution for the lower value (94.6%) was from the H&N plan for which only 85% of the points passed. Gamma results for clinical cases were: H&N (99.7%±0.9% and 97.4±7.9%), Prostate 6MV: (99.6%±1.8% and 94.7%±12.7%), Prostate 18MV results with and without the couch were nearly identical (99.8%±0.9% vs. 99.7%±1.1%, respectively). We also found the largest differences for posterior oblique beams traversing the couch rails: For 6MV prostate, there was no relative difference for beams striking the phantom only, 0.7% when treating entering through the couch, and 4.5% for beams traversing the rails.
Conclusion: Including the couch structures is important for lower energies (6MV), especially for oblique beams traversing the couch rails. However, adjusting rail positions during delivery should improve agreement for algorithms which do not account for the couch.